You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Description: The Detection Analysis Report (#60352, 442 runs / 24h window) reported average token usage as exactly 0 for both detection-enabled and regular run groups. This is not an isolated report bug: independently querying agenticworkflows logs live during this DeepReport cycle confirmed the run-metadata JSON schema it returns has no token/cost field at all — the available keys are action_minutes, actor, agent, branch, classification, created_at, display_title, duration, engine, engine_id, event, event_name, head_sha, logs_path, number, organization, ref, repository, run_attempt, run_id, sha, started_at, status, updated_at, url, workflow_name, workflow_path. Any report script that aggregates "average token usage" from this run list (rather than fetching per-run detail from logs_path) will always compute 0. This is distinct from the already-tracked chronic Daily Performance Summary 90-day-window instability issue (different symptom: metric swings, not universal zero).
Expected Impact: Fixes a systemic telemetry gap affecting every fleet-analytics report that surfaces token usage (Detection Analysis Report and likely others using the same logs summary schema), restoring a metric currently silently useless across the board.
Suggested Agent: coding agent with access to the agenticworkflows/gh-aw logs tooling; needs to determine whether token counts are tracked per-run elsewhere (e.g. only reachable via logs_path detail fetch) and either add it to the summary schema returned by logs, or fix report scripts to pull it from the correct source.
Estimated Effort: 2-4 hours (schema/tooling fix, not per-report patch).
Data Source: Detection Analysis Report discussion #60352, cross-verified live against agenticworkflows logs output during this cycle.
Warning
Firewall blocked 1 domain
The following domain was blocked by the firewall during workflow execution:
api.anthropic.com
To allow these domains, add them to the network.allowed list in your workflow frontmatter:
Description: The Detection Analysis Report (#60352, 442 runs / 24h window) reported average token usage as exactly 0 for both detection-enabled and regular run groups. This is not an isolated report bug: independently querying
agenticworkflows logslive during this DeepReport cycle confirmed the run-metadata JSON schema it returns has no token/cost field at all — the available keys areaction_minutes, actor, agent, branch, classification, created_at, display_title, duration, engine, engine_id, event, event_name, head_sha, logs_path, number, organization, ref, repository, run_attempt, run_id, sha, started_at, status, updated_at, url, workflow_name, workflow_path. Any report script that aggregates "average token usage" from this run list (rather than fetching per-run detail fromlogs_path) will always compute 0. This is distinct from the already-tracked chronic Daily Performance Summary 90-day-window instability issue (different symptom: metric swings, not universal zero).Expected Impact: Fixes a systemic telemetry gap affecting every fleet-analytics report that surfaces token usage (Detection Analysis Report and likely others using the same
logssummary schema), restoring a metric currently silently useless across the board.Suggested Agent: coding agent with access to the
agenticworkflows/gh-aw logs tooling; needs to determine whether token counts are tracked per-run elsewhere (e.g. only reachable vialogs_pathdetail fetch) and either add it to the summary schema returned bylogs, or fix report scripts to pull it from the correct source.Estimated Effort: 2-4 hours (schema/tooling fix, not per-report patch).
Data Source: Detection Analysis Report discussion #60352, cross-verified live against
agenticworkflows logsoutput during this cycle.Warning
Firewall blocked 1 domain
The following domain was blocked by the firewall during workflow execution:
api.anthropic.comTo allow these domains, add them to the
network.allowedlist in your workflow frontmatter:See Network Configuration for more information.